In recent years, studies have shown that children as young as age 3 use relational aggression in their peer relationships, that these behaviors are moderately stable, and that relational aggression is associated with concurrent and future adjustment difficulties (e.g., peer rejection). These results point to the need for investigations of developmental antecedents of relational aggression in early childhood, particularly those that focus on central contexts of development such as parent-child relationships. The few existing studies in this area have shown that broad parenting styles and practices (e.g., warmth, psychological and coercive control) are associated with relational aggression and physical aggression. Processes within the family that contribute uniquely to child relational aggression have not yet been identified. The proposed study extends prior research in several ways. First, we focus on "direct" parental influence processes, or the explicit ways in which parents meet specific socialization goals in children, in addition to broad parenting styles. In particular, we are interested in parent-child conversations about peer conflicts as an important context in which parents may communicate values and expectations, provide interpretations of social situations, and offer explicit instruction on how to interact with peers - a process called "social coaching". Second, this study will employ a short-term longitudinal design and observation methods, which will allow us to examine whether parenting behaviors predict changes in children's relational aggression and other indices of social competence over a one-year period. Finally, this study draws on social information-processing theory to make predictions about processes that contribute to parents'coaching (e.g., attributions about children) and those that might mediate the association of coaching and child relational aggression (e.g., children's normative beliefs about aggression). In the proposed study with 140 preschoolers and their primary caregivers, we will assess parents'social coaching using an adapted laboratory paradigm, and parenting styles will be assessed using observation and self-reports. We expect that the qualities of parent-child discourse about relational aggression will predict young children's use of relational aggression above and beyond the influence of broad parenting styles. A combination of multiple regression and observed variable structural equation modeling techniques will be used. This study will increase our understanding of basic processes in the development of relational aggression, in addition to informing prevention efforts targeting relational aggression in early childhood. PUBLIC HEALTH RELEVANCE: Previous research has demonstrated that relational aggression poses significant risk to children's social and psychological health and well-being, particularly among females, yet little is understood about the developmental antecedents of this behavior. This study - which focuses on processes within parent-child relationships - is expected to contribute information that will inform the development of prevention and intervention programs targeting relational aggression in early childhood.